Background: LMB-100 (previously RG7787) is a next generation recombinant immunotoxin (RIT) developed by the Pastan Lab (NCI/ LMB) in collaboration with Roche. This RIT binds to the cancer antigen mesothelin (MSLN) that is expressed on 95% of pancreatic adenocarcinomas and delivers a potent bacterial toxin to the cell cytosol. The toxin, a molecularly engineered variant of Pseudomonas exotoxin A, kills cells by irreversibly modifying a critical enzyme in the protein synthesis pathway resulting in a halt in the cell's ability to produce new proteins. This insult triggers apoptosis in many cell types. AIM 1. Determine how the combination of RG7787 and NAB-paclitaxel improves anti-tumor efficacy in pre-clinical models. We recently showed that combination of RG7787 with NAB-paclitaxel, a drug FDA approved for the treatment of pancreatic cancer, produced durable complete regressions in a mouse pancreatic cancer model and that this favorable interaction can also be modeled in cell culture. In FY 2016, we performed additional in vitro studies that enabled us to identify two potential mechanisms for this synergy. In collaboration with the Fojo lab, we found that co-administration of RG7787 with paclitaxel increased the amount of modified tubulin, a marker of paclitaxel activity. In addition, we demonstrated that RG7787 decreased levels of the survival factor mcl-1 making cells more susceptible to paclitaxel-mediated apoptosis, and that RG7787 treatment reduced levels of supportive cytokines secreted by tumor cells both in cell culture and in mouse models even at non-lethal doses. In collaboration with the Collaborative Protein Technology Resource Core, we are further examining how the drug combination changes protein levels of a broader panel of apoptosis and signaling proteins as compared to the single agents. These data were presented in poster abstract form at the AACR Molecular and Targeted Therapies meeting (11/2015, Boston) and at the AACR Special Conference on Pancreatic Cancer (5/2016, Orlando). A manuscript including results of these and prior studies from FY2015 is currently under review at Oncotarget. AIM 2. Determine whether delivery and efficacy of RG7787 to tumors can be improved by co-administration of an anti-fibrotic drug. Pancreatic adenocarcinoma contains a large volume of stroma that inhibits delivery of anti-cancer therapeutics, provides cytokines that help tumor cells survive under stress, and suppressed anti-tumor immune activation. In FY2016, we continued testing of the anti-fibrotic drug pirfenidone in collaboration with Perwez Hussain (NCI/LHC), but saw no effect on tumor stroma even in an autochthonous model of pancreatic cancer. We have since entered into a collaboration to test an anti-fibrotic/ anti-angiogenesis drug developed by an extramural investigator. We have submitted an Intramural/ Extramural Grant application to receive supplemental support for this project. AIM 3. Determine whether RG7787 or other mesothelin-targeted RITs can boost the effect of immune activating drugs like checkpoint inhibitors. Pancreatic adenocarcinoma produces an immunosuppressive microenvironment. Killing tumor cells with oncolytic viruses or administering anti-tumor vaccines can cause immune activation. Combining these treatments with immune checkpoint inhibition has been demonstrated to produce anti-tumor immune responses in pre-clinical models of pancreatic cancer. We hypothesize that since RG7787 uses a bacterial toxin to kill tumor cells, it may also induce immune activation within the pancreatic cancer microenvironment that could be leveraged to induce an anti-tumor immune response in combination with immune checkpoint inhibition. Because our immunotoxins bind only to human (hMSLN) and not to native mouse MSLN (mMSLN), completing this project requires introduction of a syngeneic mouse pancreatic cancer cell line expressing hMSLN into a mouse strain that will not reject these cells. In FY2016, we created two Panc02 cell lines that stably express hMSLN. To overcome this issue, we initiated a collaboration with the CAPR group at Frederick (Serguei Kozlov, Leidos) to develop two new syngeneic mouse models limited expression of hMSLN. Founders have been identified in both of these models, and breeding continues in order to identify useful lines AIM 4 (continuing). Evaluate efficacy of MSLN-targeted RITs in pancreatic cancer patients. At least 70% of pancreatic adenocarcinomas express the surface antigen MSLN, making these tumors good targets for MSLN-targeted therapies. Raffit Hassan (NCI/ TGIB) demonstrated that the MSLN-targeted RIT SS1P can produce durable major responses in heavily pre-treated patients with malignant mesothelioma (Hassan et al, Sci Trans Med, 2013), but this combination had not yet been tested in pancreatic cancer. In collaboration with Dr. Hassan, his clinical protocol testing the SS1P with pentostatin + cyclophosphamide regimen was amended to include pancreatic cancer patients. Four pancreatic cancer patients have been treated on this study. No clinical responses were observed. Based on our results from AIM1, we recently initiated A Phase Ib/II Study of Mesothelin-Targeted Immunotoxin LMB-100 in Combination with Nab-Paclitaxel in Participants with Previously Treated Metastatic and/ or Locally Advanced Pancreatic Ductal Adenocarcinoma (PI Alewine) in order to determine the safety, tolerability and efficacy of the LMB-100 + NAB-paclitaxel combination. This trial received IRB approval in 6/2016 and formally opened for accrual 8/2016.